62 S. L. Dingman et al. 



L = empirical constant = maximum thaw depth (35 cm in 1970; 

 37.5 cm in 1971; 63.3 cm in 1972). 



Note that there is considerable year-to-year variation in the thaw pro- 

 gression; this variation can be correlated with cumulative net radiation 

 during the early summer. There is also marked spatial variation in thaw 

 progression due to soil type. 



As the summer season begins in late June or early July, net radiation 

 decreases. This is due to the passing of the summer solstice, an increase 

 in cloudiness (Figure 2-9), and an increase of the albedo of the tundra to 

 an average of about 19% as the surface dries. Evaporation still consumes 

 the largest portion of the available energy, but convective heat loss in- 

 creases in importance (Figure 2-8). Diurnal variations in soil temperature 

 are greatest during this period as a result of strong diurnal changes in ra- 

 diation under snow-free conditions (Kelley and Weaver 1969). 



Summer soil temperatures vary across microtopographic positions 

 and during the summer these differences reflect variations in albedo, mi- 

 croclimate and soil properties, particularly those related to moisture con- 

 tent. Elevated rims of low-centered polygons are sometimes cooler due to 

 wind, while troughs and basins are warmer. However, at other times, in- 

 creased evaporation and transpiration from the more vegetated and/or 

 wetter troughs may result in lower temperatures there (Figure 2-14). 

 Maximum differences occur under clear sky conditions during early af- 

 ternoon with 8°C differences observed at the 1-cm depth between cooler 

 rims and warmer basins and troughs of low-centered polygons (Goodwin 

 1976). Nighttime difference decreases 2 to 3°C. Although diurnal and 

 seasonal soil temperatures follow closely changes in air temperature, 

 other climatic factors such as cloud cover modify the magnitude of the 

 difference between them. For example, soil temperatures at 1 cm depth 

 for rims, troughs and basins at site 4 averaged 8.7 °C in July 1972 and 

 5.6°C in July 1973. Average monthly air temperatures for July 1972 were 

 only 1.8°C higher than 1973 (Table 2-1). Increased radiational warming 

 accounted for most of the increased soil temperature. 



Evapotranspiration rates decrease from the post-melt season be- 

 cause of the decrease in water on the surface and the decrease in available 

 energy (Weller and Holmgren 1974a). Studies have consistently shown a 

 near balance of precipitation with evapotranspiration during the summer 

 (Mather and Thornthwaite 1958, Brown et al. 1968, Guymon 1976). 

 About 80% of the annual evapotranspiration occurs during the 1 July-31 

 August season. Comparisons of total evapotranspiration and pan evap- 

 oration indicate that water losses from moist vegetated surfaces are ap- 

 proximately the same as those from open water. Koranda et al. (1978) 

 measured overall loss rates from soil as 4.6 to 5.6 mm day', compared 

 with open-pan evaporation of 3 mm day '. Based on Miller et al. (1976) 



